What's Changed
Breaking Changes 🛠
Bug fixes 🐛
- fix(parler-tts): use latest audiotools by @mudler in #3954
- fix(parler-tts): pin grpcio-tools by @mudler in #3960
- fix(gallery): overrides for parler-tts in the gallery by @mudler in #3962
- fix(grpc): pass by modelpath by @mudler in #4023
- fix(tts): correctly pass backend config when generating model options by @mudler in #4091
Exciting New Features 🎉
- feat(vllm): expose 'load_format' by @mudler in #3943
- feat(tts): Implement naive response_format for tts endpoint by @n-Arno in #4035
- feat(diffusers): allow multiple lora adapters by @mudler in #4081
- feat(ui): move model detailed info to a modal by @mudler in #4086
- feat: allow to disable '/metrics' endpoints for local stats by @mudler in #3945
- feat: add flux single file support by @sozercan in #3959
🧠 Models
- fix(phi3-vision): add multimodal template by @mudler in #3944
- models(gallery): add l3.1-moe-2x8b-v0.2 by @mudler in #3969
- models(gallery): add llama3.1-darkstorm-aspire-8b by @mudler in #3970
- models(gallery): add llama-3.2-sun-2.5b-chat by @mudler in #3971
- models(gallery): add darkest-muse-v1 by @mudler in #3973
- models(gallery): add llama-3.2-3b-instruct-uncensored by @mudler in #3974
- models(gallery): add thebeagle-v2beta-32b-mgs by @mudler in #3975
- models(gallery): add l3.1-70blivion-v0.1-rc1-70b-i1 by @mudler in #3977
- models(gallery): add llama-3.1-hawkish-8b by @mudler in #3978
- models(gallery): add quill-v1 by @mudler in #3980
- models(gallery): add delirium-v1 by @mudler in #3981
- models(gallery): add magnum-v4-9b by @mudler in #3983
- models(gallery): add llama-3-whiterabbitneo-8b-v2.0 by @mudler in #3984
- models(gallery): add l3-nymeria-maid-8b by @mudler in #3985
- models(gallery): add meraj-mini by @mudler in #3987
- models(gallery): add granite-3.0-1b-a400m-instruct by @mudler in #3994
- models(gallery): add moe-girl-800ma-3bt by @mudler in #3995
- models(gallery): add spiral-da-hyah-qwen2.5-72b-i1 by @mudler in #4022
- models(gallery): add llama3.1-bestmix-chem-einstein-8b by @mudler in #4028
- models(gallery): add starcannon-unleashed-12b-v1.0 by @mudler in #4032
- models(gallery): add smollm2-1.7b-instruct by @mudler in #4033
- models(gallery): add control-8b-v1.1 by @mudler in #4039
- models(gallery): add whiterabbitneo-2.5-qwen-2.5-coder-7b by @mudler in #4042
- models(gallery): add llama-3.1-whiterabbitneo-2-8b by @mudler in #4043
- models(gallery): add g2-9b-aletheia-v1 by @mudler in #4056
- models(gallery): add cybertron-v4-qw7b-mgs by @mudler in #4063
- models(gallery): add g2-9b-sugarquill-v0 by @mudler in #4073
- chore(model-gallery): ⬆️ update checksum by @localai-bot in #4080
- models(gallery): add q25-1.5b-veolu by @mudler in #4088
- models(gallery): add valor-7b-v0.1 by @mudler in #4089
- models(gallery): add tess-r1-limerick-llama-3.1-70b by @mudler in #4095
- models(gallery): add llenn-v0.75-qwen2.5-72b-i1 by @mudler in #4098
- models(gallery): add eva-qwen2.5-14b-v0.2 by @mudler in #4099
- models(gallery): add opencoder-8b instruct and base by @mudler in #4101
- models(gallery): add opencoder-1.5b instruct and base by @mudler in #4102
- models(gallery): add tissint-14b-128k-rp by @mudler in #4103
- models(gallery): add tq2.5-14b-sugarquill-v1 by @mudler in #4104
📖 Documentation and examples
- docs: add Homebrew as an option to install on MacOS by @mauromorales in #3946
- chore: drop examples folder now that LocalAI-examples has been created by @dave-gray101 in #4017
👒 Dependencies
- chore(deps): Bump llama-index from 0.11.17 to 0.11.19 in /examples/chainlit by @dependabot in #3893
- chore(deps): Bump weaviate-client from 4.8.1 to 4.9.0 in /examples/chainlit by @dependabot in #3894
- chore(deps): Bump langchain from 0.3.3 to 0.3.4 in /examples/functions by @dependabot in #3900
- chore(deps): Bump langchain-community from 0.3.2 to 0.3.3 in /examples/langchain/langchainpy-localai-example by @dependabot in #3923
- chore(deps): Bump docs/themes/hugo-theme-relearn from
007cc20to06e70daby @dependabot in #3932 - chore(deps): Bump sqlalchemy from 2.0.35 to 2.0.36 in /examples/langchain/langchainpy-localai-example by @dependabot in #3920
- chore(deps): Bump yarl from 1.15.2 to 1.15.5 in /examples/langchain/langchainpy-localai-example by @dependabot in #3921
- chore(deps): Bump openai from 1.51.2 to 1.52.0 in /examples/langchain-chroma by @dependabot in #3908
- chore(deps): Bump llama-index from 0.11.17 to 0.11.19 in /examples/langchain-chroma by @dependabot in #3907
- chore(deps): Bump marshmallow from 3.22.0 to 3.23.0 in /examples/langchain/langchainpy-localai-example by @dependabot in #3917
- chore(deps): Bump openai from 1.51.2 to 1.52.0 in /examples/functions by @dependabot in #3901
- chore(deps): Bump yarl from 1.15.5 to 1.16.0 in /examples/langchain/langchainpy-localai-example by @dependabot in #3938
- chore(deps): Bump openai from 1.51.2 to 1.52.2 in /examples/langchain/langchainpy-localai-example by @dependabot in #3993
- chore(deps): Bump torchvision from 0.18.1+rocm6.0 to 0.20.0+cu118 in /backend/python/diffusers by @dependabot in #3997
- chore(deps): Bump docs/themes/hugo-theme-relearn from
06e70dato28fce6bby @dependabot in #3986 - chore(deps): Bump llama-index from 0.11.19 to 0.11.20 in /examples/langchain-chroma by @dependabot in #3988
- chore(deps): Bump openai from 1.52.0 to 1.52.2 in /examples/langchain-chroma by @dependabot in #3989
- chore(deps): Bump llama-index from 0.11.19 to 0.11.20 in /examples/chainlit by @dependabot in #3990
- chore(deps): Bump tqdm from 4.66.5 to 4.66.6 in /examples/langchain/langchainpy-localai-example by @dependabot in #3991
- chore(deps): Bump frozenlist from 1.4.1 to 1.5.0 in /examples/langchain/langchainpy-localai-example by @dependabot in #3992
- chore(deps): Bump openai from 1.52.0 to 1.52.2 in /examples/functions by @dependabot in #4000
- chore(deps): bump grpcio to 1.67.1 by @mudler in #4009
- chore(deps): bump llama-cpp to 8f275a7c4593aa34147595a90282cf950a853690 by @mudler in #4016
Other Changes
- docs: ⬆️ update docs version mudler/LocalAI by @localai-bot in #3936
- chore: ⬆️ Update ggerganov/llama.cpp to
e01c67affe450638162a1a457e2e57859ef6ebf0by @localai-bot in #3937 - chore: update go-piper to latest by @dave-gray101 in #3939
- chore: ⬆️ Update ggerganov/llama.cpp to
c8c07d658a6cefc5a50cfdf6be7d726503612303by @localai-bot in #3940 - chore: ⬆️ Update ggerganov/whisper.cpp to
0fbaac9c891055796456df7b9122a70c220f9ca1by @localai-bot in #3950 - chore: ⬆️ Update ggerganov/llama.cpp to
0a1c750c80147687df267114c81956757cc14382by @localai-bot in #3948 - chore: ⬆️ Update ggerganov/llama.cpp to
958367bf530d943a902afa1ce1c342476098576bby @localai-bot in #3956 - chore(python): update backend sample to consume grpcio from venv by @mudler in #3961
- fix(ci): drop grpcio-tools pin to apple CI test run by @mudler in #3964
- chore: ⬆️ Update ggerganov/llama.cpp to
668750357e66bfa3d1504b65699f5a0dfe3cb7cbby @localai-bot in #3965 - fix(parler-tts): pin protobuf by @mudler in #3963
- chore: change url to github repository by @mudler in #3972
- chore: ⬆️ Update ggerganov/llama.cpp to
cc2983d3753c94a630ca7257723914d4c4f6122bby @localai-bot in #3976 - chore: ⬆️ Update ggerganov/llama.cpp to
8841ce3f439de6e770f70319b7e08b6613197ea7by @localai-bot in #3979 - Revert "chore(deps): Bump torchvision from 0.18.1+rocm6.0 to 0.20.0+cu118 in /backend/python/diffusers" by @mudler in #4008
- chore: ⬆️ Update ggerganov/llama.cpp to
61715d5cc83a28181df6a641846e4f6a740f3c74by @localai-bot in #4006 - chore: ⬆️ Update ggerganov/whisper.cpp to
d4bc413505b2fba98dffbb9a176ddd1b165941d0by @localai-bot in #4005 - chore: ⬆️ Update ggerganov/whisper.cpp to
55e422109b3504d1a824935cc2681ada7ee9fd38by @localai-bot in #4015 - chore(tests): expand timeout by @mudler in #4019
- chore: ⬆️ Update ggerganov/llama.cpp to
61408e7fad082dc44a11c8a9f1398da4837aad44by @localai-bot in #4021 - chore: ⬆️ Update ggerganov/whisper.cpp to
19dca2bb1464326587cbeb7af00f93c4a59b01fdby @localai-bot in #4020 - chore(tests): bump timeouts by @mudler in #4024
- chore: ⬆️ Update ggerganov/llama.cpp to
ab3d71f97f5b2915a229099777af00d3eada1d24by @localai-bot in #4025 - chore: ⬆️ Update ggerganov/whisper.cpp to
aa037a60f32018f32e54be3531ec6cc7802899ebby @localai-bot in #4026 - chore: ⬆️ Update ggerganov/whisper.cpp to
0377596b77a3602e36430320cbe45f8c305ef04aby @localai-bot in #4031 - chore: ⬆️ Update ggerganov/llama.cpp to
418f5eef262cea07c2af4f45ee6a88d882221fcbby @localai-bot in #4030 - chore(docs): add top-header partial by @mudler in #4034
- chore: ⬆️ Update ggerganov/llama.cpp to
42cadc74bda60afafb45b71b1a39d150ede0ed4dby @localai-bot in #4037 - docs: Update documentation for text-to-audio feature regarding response_format by @n-Arno in #4038
- chore: ⬆️ Update ggerganov/llama.cpp to
9830b6923b61f1e652a35afeac77aa5f886dad09by @localai-bot in #4040 - chore: ⬆️ Update mudler/go-piper to
e10ca041a885d4a8f3871d52924b47792d5e5aa0by @localai-bot in #3949 - chore: ⬆️ Update ggerganov/llama.cpp to
9f409893519b4a6def46ef80cd6f5d05ac0fb157by @localai-bot in #4041 - chore: ⬆️ Update ggerganov/llama.cpp to
d5a409e57fe8bd24fef597ab8a31110d390a6392by @localai-bot in #4079 - chore: ⬆️ Update ggerganov/llama.cpp to
b8deef0ec0af5febac1d2cfd9119ff330ed0b762by @localai-bot in #4083 - fix(diffusers): correctly parse height and width request without parametrization by @mudler in #4082
- chore: hide raw safetensors files by @mudler in #4085
- chore: ⬆️ Update ggerganov/llama.cpp to
5c333e014059122245c318e7ed4ec27d1085573cby @localai-bot in #4087 - chore(p2p): enhance logging by @mudler in #4090
- chore: ⬆️ Update ggerganov/llama.cpp to
97404c4a0374cac45c8c34a32d13819de1dd023dby @localai-bot in #4093 - chore: ⬆️ Update ggerganov/whisper.cpp to
31aea563a83803c710691fed3e8d700e06ae6788by @localai-bot in #4092 - fix(container-images): install uv as system package by @mudler in #4094
- chore(refactor): drop unnecessary code in loader by @mudler in #4096
- chore: ⬆️ Update ggerganov/llama.cpp to
ec450d3bbf9fdb3cd06b27c00c684fd1861cb0cfby @localai-bot in #4097 - chore: ⬆️ Update ggerganov/llama.cpp to
6423c65aa8be1b98f990cf207422505ac5a441a1by @localai-bot in #4100
New Contributors
Full Changelog: v2.22.1...v2.23.0